![]() | Only 14 pages are availabe for public view |
Abstract Despite conciderable research activity in the area of structural optimization during the last twenty five years, and despite the new and improved algorithms that emerged during this period, structural optimization algorithms have not been applied exensively to practical real-world problems. The mathematical optimization of a realistic structure subjected to the constrains of actual design specifications is often highly complex and involves a large number of design variables and constraints, thus requiring unacceptably large computer processing time. This is particulary true in the case of optimal desig of structures subjected to dynamic loads due to the presence of the time parameter. After decades of using and improving single processor computer technology and developing computer programs that tackle problems in a serial fashion, researchers have realized that they arefinally reaching the limite. Copmuter chips cannot get much smaller and electricity cannot travel faster. In the quest for overcoming such physical limits, the idea of multiprocessor computers emerged. Various types of such multiprocessor computers have recently become commercially available. For a survey of some of these computers, see Adeli and Vishnuhotla (1987). In such computers, several processors jointly tackle the same problem at the same time. T he challenge now is to device new algorithms and stratagems in order to achive maximum utilization of the concurrent processing capabilties of this new generation of computers. |